EP4033090A1 - Method for controlling a wind energy system - Google Patents

Abstract

Method for controlling a wind power plant (100), the wind power plant (100) having an aerodynamic rotor (106) which is operated at a variable speed and which has rotor blades (108) with an adjustable blade angle, the wind power plant (100) in a partial load range is controlled by an operating characteristic control that uses an operating characteristic, the operating characteristic (402) specifying a relationship between the rotational speed and a generator state variable to be set, the generator state variable to be set being a generator output or a generator torque, and the operating characteristic control takes place in such a way that depending on a detected speed, a value of the generator state variable predetermined by the operating characteristic is set, and the wind turbine is controlled in a full-load range by pitch control (304), in which the speed is adjusted to a speed setpoint by adjusting the blade angle rt is controlled, with the wind turbine being controlled by a speed power controller (302) in at least one predefinable speed range of the partial load range and/or in a transition range from the partial load range to the full load range, in which the speed is controlled by adjusting the generator state variable to a speed setpoint, the Speed power control takes place in such a way that in an outer cascade from a comparison of a specified desired speed with a detected actual speed via a first controller (310), a first acceleration setpoint (a _s ) of the rotor is determined, in an inner cascade from a comparison of the first acceleration desired value with a detected actual acceleration value of the rotor via a second controller (316), a generator setpoint is determined as a setpoint for the generator state variable.

Description

The present invention relates to a method for controlling a wind power plant. The present invention also relates to a corresponding wind energy installation.

Wind turbines are known, they generate electrical power from the wind. For this purpose, they usually have a rotor with rotor blades that are moved by the wind. The rotor then rotates at a rotor speed that also depends on the wind speed and drives a generator.

To control this power generation, the rotor blades can be adjusted in their angle of attack, which is referred to as pitching. In addition, the power output or a generator torque of the generator can be influenced to control the wind energy installation. This also results in a variable rotor speed and thus a variable generator speed. In the case of a gearless wind turbine, the rotor speed corresponds to the generator speed.

The control of the wind energy installation has the particular task of ensuring the best possible performance-optimal operation at low speeds, in which as much power as possible is generated. At high wind speeds, the wind turbine is closed control that a speed limit and a power limit are observed.

At low wind speeds, one can also speak of part-load operation or part-load range and at high wind speeds of full-load operation or full-load range. In part-load operation, as much power as possible is to be generated, and in full-load operation the wind energy installation is to be controlled in such a way that it is protected against overload. Corresponding control concepts are known. Accordingly, a characteristic curve control is usually used for part-load operation, in which an operating characteristic curve specifies a relationship between the rotational speed and the power to be set for it. The blade angle is usually constant. In full-load operation, speed control is usually used, which tries to keep the speed constant with constant power by adjusting the rotor blades.

Additional special features can be added to these basic requirements. It often happens that the speed range in part-load operation also includes a speed range in which a resonance point of the wind energy installation is located. In order to circumvent this problem, it can be provided that such a resonance point is bypassed by avoiding speeds that are in such a resonance range. However, bypassing a speed in this way can be problematic, because this results in a point of discontinuity in the operating characteristic. This can lead to problems in the practical implementation of such a reversal.

Further problems can arise in the transition range from partial load operation to full load operation. It is particularly critical here that the operating characteristic is comparatively steep, so that a small change in speed leads to a large change in power, which can be a challenge in terms of control technology and can cause problems similar to a point of discontinuity. In addition, fluctuations in the wind speed for such discontinuities or almost discontinuous areas can lead to high, possibly too high, controller activity.

The object of the present invention is to address at least one of the problems mentioned above. In particular, a solution is to be proposed in which a control is described that also works well in speed ranges that are in the range of a resonant frequency and/or in the range of a transition from part-load operation to full-load operation. At the very least, an alternative solution to previously known solutions should be proposed.

According to the invention a method according to claim 1 is proposed. This method for controlling a wind power plant is based on a wind power plant that has an aerodynamic rotor that is operated at a variable speed and that has rotor blades with an adjustable blade angle. The wind energy installation can be operated in a partial load range and in a full load range. In the part-load range, wind speeds up to a nominal wind speed are present, and in the full-load range, wind speeds above the nominal wind speed are present. The areas can easily overlap, especially since the wind speed can fluctuate and cannot be determined with any degree of accuracy.

In this case, the wind energy installation is controlled in the partial load range by an operating characteristic control which uses an operating characteristic. The operating characteristic specifies a relationship between the speed, namely the rotor speed, and a generator state variable to be set. The generator state variable to be set can be a generator power or a generator torque. The operating characteristic curve is controlled in such a way that a value of the generator state variable that is predetermined by the operating characteristic curve is set as a function of a detected rotational speed. An operating characteristic can also be specified by a functional description, e.g. a formula or function. The operating characteristic can also be stored as a table.

If the generator state variable is the generator power, then a speed is recorded and the operating characteristic shows an associated value of the generator power that is to be set for the recorded speed. This generator output is then adjusted. Usually the blade angle is left unchanged. Depending on the rotational speed, the set generator power then leads to a generator torque that counteracts a rotor torque that is generated in the rotor by the wind.

If these two torques are the same, a stable operating point has resulted. If these two torques are not equal, the rotor will speed up or slow down. The rotor speed therefore changes and a new value for the generator power is determined from the operating characteristic and set until a stable operating point is obtained.

In the full-load range, the wind energy installation is controlled by pitch control, in which the speed is controlled by adjusting the blade angle to a desired speed value. In particular, the speed is regulated here to a nominal speed value. This serves in particular to protect the wind turbine from overloading. So if the speed rises above this speed value, to which the wind energy installation is to be regulated the rotor blades are adjusted in their blade angle, namely turned out of the wind, at least by one or a few degrees. As a result, less power is taken from the wind and the speed can be reduced or maintained at the desired value.

It is further proposed that in at least one predefinable speed range of the part-load range and/or in a transition range from the part-load range to the full-load range, the wind turbine is controlled by speed power control, in which the speed is controlled by adjusting the generator state variable to a speed setpoint. In these areas, there is a deviation from the operating characteristic curve control otherwise provided for the partial load range.

This can be provided in particular for a speed range of the partial load range, which thus forms a small section of the partial load range in which a resonance frequency can occur or in which a resonance point is present. In particular, this predefinable speed range of the part-load range can be referred to as the avoidance range, in which there is a speed that should be avoided. The rotational speed can then be regulated to a rotational speed at the beginning or at the end of the predefinable rotational speed range of the part-load range by means of this rotational speed output control provided. In this way, the speed that is in particular in the middle of such a speed range, that is to say in the middle of the avoidance range, can be left out.

Two control cascades are provided for this speed power control. The rotational speed output is therefore such that a first setpoint acceleration value of the rotor is determined in an outer cascade from a comparison of a predefined setpoint rotational speed with a detected actual rotational speed via a first controller. To this extent, this speed control error is transferred to this acceleration setpoint. This desired acceleration value of the rotor can thus be understood as an acceleration value by which the rotor should accelerate. The acceleration setpoint can also assume negative values and indicate how the speed should decrease.

It is then further provided that a generator setpoint is determined as a setpoint for the generator state variable in an internal cascade from a comparison of the first acceleration setpoint with a detected actual acceleration value of the rotor via a second controller. The second controller thus sets the control error of the rotor acceleration in a generator power setpoint or a generator torque setpoint depending on which of the two variants is provided for the generator state variable.

Thus, for the operating characteristic curve control in the partial load range, a change to a speed control is proposed for special areas. It is therefore changed from a control to a regulation. A special cascade structure is provided for the control, namely the speed control, which controls the speed in the outer cascade and controls a torque in the inner cascade.

In addition, it was recognized in particular that the torque and thus also the power can be well controlled by such a speed control at points of discontinuity or points with a very large gradient in a speed-power characteristic. In particular, it was recognized here that large changes in performance can occur with small changes in speed, which can be controlled in this way. It can be avoided that large rotor accelerations occur. In addition, this enables or simplifies an interaction between the speed controller and a power controller. This is explained below in at least one exemplary embodiment.

According to one aspect, it is proposed that the speed power control has an integral component, in particular that the second controller has an integral component. Here it was particularly recognized that in the case of discontinuities or very steep speed/power curves, it can make sense to keep the speed at a value for which the integral component can be used. The power output can also be maintained without permanent control errors thanks to the integral component.

In particular, it was recognized that it is not necessary to provide this integral component in the outer controller cascade. This avoids a situation where a discrepancy between the setpoint speed and the actual speed leads to an increasing value of the setpoint torque. Instead, the integral term can be used in the inner cascade. The inner cascade generates a generator output or a generator torque from the setpoint/actual value comparison for the acceleration value, which in this respect can be viewed as a manipulated variable. A run-up of this manipulated variable by the integral component is thus immediately implemented in the generator and could immediately take effect, so that the integral component is proposed in this inner cascade.

According to one aspect, it is proposed that the first acceleration setpoint and the actual acceleration value are each designed as an acceleration power in the rotational speed power control. The acceleration power is associated with a rotor acceleration and describes a power required to produce the rotor acceleration. As a result, the difference between a target and actual acceleration in particular can also be expressed and evaluated as a performance. Such a power can also be specifically related to a power that is available for regulation. It is particularly advantageous to facilitate a comparison with the regulation in the full-load range. Both regulations can be compared by considering this performance in corresponding manipulated variables or control variables.

According to one aspect, it is proposed that in the rotational speed power control, the outer cascade for determining the first desired acceleration value has at least one acceleration limit value. If an acceleration power is considered for the acceleration, corresponding power limit values can be provided. The speed control can thus be kept within relevant limits. In particular, it is possible to take into account how much power is actually available for implementing the speed control. This advantage is not only achievable when an acceleration performance is considered for the acceleration. Both the power available from the wind and the adjustable generator power can be taken into account as the available power.

Here it can be particularly taken into account that the maximum power for driving the rotor is given by the wind speed. At a certain working point, a large part of the available power from the wind should already be used. A differential power can then remain up to the maximum power that can be drawn, which is available for the speed control. This can be taken into account in the at least one acceleration limit value. The situation is similar with the generator, which also cannot be adjusted at will from its current operating point and can therefore only provide power for the speed control up to a certain value. That can be taken into account. A lower value can also be taken into account when reducing the power drawn from the wind at an operating point. Such a lower value can—to exaggerate it somewhat graphically—be one at which the wind energy installation threatens to come to a standstill.

Accordingly, it is preferably proposed that the at least one acceleration limit value can be set. It can thereby be adapted to the circumstances explained.

In addition or as an alternative, it is proposed that an upper and a lower acceleration limit value with different values be provided as the at least one acceleration limit value. It is also particularly taken into account here that, for aerodynamic reasons, the power available for increasing the speed can be of a different size than the power for reducing the speed.

In addition or as an alternative, it is proposed that the inner cascade for determining the manipulated variable for setting the generator state variable has an integral element with an integrator limitation. Such an integrator limitation can optionally be set and, in addition or as an alternative, different upper and lower limit values are provided for the integrator limitation. Advantages of using the integral term in the inner cascade and thus in the second controller have already been described above.

It is also proposed here to provide an integrator limitation for the respective integral element. In this way, it can be prevented in particular that the integral component continues to rise or continues to fall in the negative direction in the event of a remaining control deviation. Here, too, it was taken into account that only a limited amount of power is available as actuating energy for implementing the speed control. By limiting the integral element, this is taken into account and it can thereby be avoided that the integral element specifies a manipulated variable, in particular a corresponding setpoint power, which cannot be implemented at all. In order to take this into account particularly well, it is proposed that the integrator limitation be preset in an adjustable manner and/or that corresponding power limits be taken into account by means of different upper and lower limit values.

According to one aspect, it is proposed that the at least one predefinable speed range is provided as a speed avoidance range, which is characterized by a lower avoidance speed and an upper avoidance speed. For this purpose, it is provided that the speed power control is used when the lower avoidance speed is reached with increasing speed or the upper avoidance speed is reached with falling speed. In particular, it is provided that such an avoidance speed range is formed around a speed that forms a resonant frequency or contributes to exciting a resonance in the wind energy installation.

In addition, it is proposed that when using speed power control, the wind turbine can be operated at a lower operating point, depending on a prevailing wind speed, in a lower speed range, which includes the lower avoidance speed, and in an upper speed range, which includes the upper avoidance speed, in one upper operating point is operable, and a change between the lower and the upper operating point is performed depending on an aerodynamic evaluation of the lower and upper operating point.

This aspect is therefore based on the assumption that, when using the speed power control, the wind energy installation can be operated in one of these two working points, namely basically below or above the resonance situation to be avoided, to put it clearly. If the wind speed then changes, a change from the upper operating point to the lower operating point or vice versa can be considered. For such a change, it is proposed to carry out an aerodynamic evaluation of the lower and upper working point.

This is based in particular on the idea that by using the speed control, which is carried out here by the proposed speed power control, the speed is essentially constant despite changing wind speeds. The speed is therefore not well suited as a criterion as to whether a change between the two operating points makes sense. It is therefore proposed to carry out an aerodynamic evaluation of the two working points and in particular to change them when the other working point is aerodynamically better or is more suitable.

Here it was also recognized in particular that a pure assessment of the generator power, which is adjusted with this speed power control, can be unsuitable.

According to one aspect, it is proposed that an at least two-stage test be carried out to switch between the lower and upper working point. In a first stage, it is checked whether the working point to which the change is to be made has a higher aerodynamic efficiency than the working point from which the change is to be made. In a second stage, it is checked whether the generator state variable, which is determined with the part-load speed control, has reached an upper or lower generator state limit. Such a generator state limit can in particular be a predefined maximum power value. If the generator power reaches this value, the generator power increases up to this value when operating at the lower operating point on, this leads to a change from the lower to the upper working point. Accordingly, a lower power value can be specified, which leads to a change from the upper to the lower operating point when the generator power falls to this lower power limit.

In particular, it is envisaged that a change will be permitted depending on the test in the first stage, and a change will be prescribed depending on the test in the second stage. If the current working point is less favorable than the other working point, i.e. if it has a lower aerodynamic efficiency than the working point to which the change is to be made, a change can be considered. However, the change does not have to be carried out in this case and can rather depend on other conditions. Such a different condition can be, for example, how much the wind varies. If it fluctuates strongly, a change between the working points is made more cautiously. However, if the wind speed hardly fluctuates, it may make more sense to switch to the working point with the higher aerodynamic efficiency, since it can be assumed that this will remain the working point with the higher aerodynamic efficiency for a while.

However, if a limit is reached, which is determined in the second stage, a change must be made. This is based in particular on the knowledge that non-optimal operating points can be tolerated for a while. The disadvantages can be weighed against other criteria.

For example, an operating point with a higher aerodynamic efficiency promises a higher yield. However, this can be compared with the loss of yield due to the change and then it can also be taken into account whether frequent changes are to be expected or not.

However, it was also recognized that an operating point that is too unfavorable can possibly jeopardize the operation of the wind energy installation as a whole. For example, if the wind turbine is operated in the lower speed range, in particular at the lower avoidance speed, with ever-increasing power, this can result in the working point becoming so aerodynamically unfavorable that the wind turbine stops, to name just one example.

According to one aspect, it is proposed that a change between the lower and upper working point is carried out when the working point to which the change is made is to have a higher aerodynamic efficiency for a predeterminable test period than the working point from which the change is to be made. In this way it can be avoided that, despite the higher aerodynamic efficiency of the other working point, a premature change is made. In particular, the predeterminable test period is in a range from 5 seconds to 10 minutes, in particular in the range from 10 seconds to 5 minutes. These periods of time have a lower value that can be suitable for checking for a sufficiently long time whether the operating points or the wind conditions on which they are based are stable. The two upper values of 5 and 10 minutes indicate a time in which exchange losses no longer predominate. The specific choice can particularly depend on how much the two operating points differ. This in turn can depend on how large the speed avoidance range must be selected.

According to one aspect, it is proposed that the speed setpoint be specified as a constant for the speed power control or is specified via at least one speed characteristic, the speed characteristic having speed values in each case as a function of the generator state variable to be set. In particular, it is provided that the speed characteristic curve is designed in such a way that it has a negative gradient at least in sections, so that the speed values decrease with increasing values of the generator state variable.

Thus, a steep, in particular vertical, branch is suggested according to a speed/power diagram. The explanations apply analogously to the generator torque as a generator state variable if a speed/torque diagram is used. The fact that the speed setpoint is specified as a constant means that when the wind increases, an increase in the speed is countered by a corresponding increase in power. A steady-state accuracy can be achieved for the specified constant speed reference value in particular by a controller with an integral part, namely by a correspondingly high generator power as the manipulated variable. As the wind increases, the power increases without the speed increasing. Accordingly, when the wind drops, the speed can remain constant, while the generator power decreases as a manipulated variable.

It has been recognized that a negative slope can even have advantages over the vertical slope. A negative gradient is therefore a constellation in which, with increasing wind speed, the speed does not increase due to the speed power control and even decreases somewhat, while the generator power as a manipulated variable increases significantly. For example, the generator output can increase by 10%, while the speed drops again by 1%. Such a regulation can prevent an ambiguous speed-power branch from occurring in the speed-power representation. In other words, a clear performance value can be assigned to each speed on this branch.

Correspondingly, when the wind speed falls, when the wind turbine has reached the upper operating point, the generator output can continue to decrease as a manipulated variable, but the speed can even increase slightly again. Here, too, an example that can be considered is that the generator output decreases by 10% while the speed only decreases by 1%. A corresponding straight line or linear relationship can be defined as a result. In other words, a power reduction of 5% would be associated with a speed increase of 0.5%, etc. The same applies analogously to an increase in wind speed and thus an increase in power starting from the lower operating point.

According to one aspect, it is proposed that a changeover time that is preferably less than 20 seconds, in particular less than 10 seconds, is specified for changing between the lower and upper working point. In addition or as an alternative, it is proposed that a time profile for the generator state variable to be set be specified in order to thereby control the change.

The wind energy installation is operated and maintained at an operating point, that is to say at an approximately predetermined speed, as a result of the speed power control. For this purpose, the generator power is set accordingly as a generator state variable and is constantly readjusted. Of course, all of this can also be carried out using the generator torque as a generator state variable.

If you now want to switch between the operating points, for example from the lower operating point to the upper operating point, then this generator state variable, which therefore forms the manipulated variable for speed power control, is withdrawn. This allows the rotor to accelerate. The more this generator state variable is reduced for the change from the lower to the upper operating point, the faster the change can be completed. The changeover time can therefore be realized via this.

When choosing the changeover time, boundary conditions should preferably also be taken into account, so that an arbitrarily fast changeover does not have to be the only sensible goal. Limit values must also be taken into account. In particular, it should be taken into account with which value is to be calculated for the generator state variable after the change. For example, the value of the generator state variable to be expected after the change could also be set for the change process. After the change, this operating point is then controlled directly with the associated generator state variable.

However, this can lead to a too slow change. The generator state variable would therefore have to be set correspondingly lower when changing from the lower to the upper operating point. If it is set too low in order to change too quickly, an overshoot could result from the fact that the generator state variable does not reach the desired value quickly enough when the desired speed is reached after the change. Accordingly, it is proposed to specify a changeover time and the generator state variable to be set can be determined, for example, in a simulation, or from empirical values or specific tests.

The same procedure is followed accordingly when changing from the upper to the lower operating point, at which the value of the generator state variable is increased accordingly, but should not be increased too much. A suitable value for the generator state variable can also be determined for this via the proposed specification of the changeover time.

It is therefore additionally or alternatively proposed to specify a time profile for the generator state variable to be set. For the change from the lower to the upper operating point, such a time course can appear in particular in such a way that the generator state variable falls continuously for a certain time, in particular for half the changeover time, and rises continuously for the remaining time, namely in particular to the value for the controlled upper working point was determined as reasonable, for example calculated or determined by a simulation.

Correspondingly, the generator state variable would correspondingly first increase and then decrease if a change is to be made from the upper operating point to the lower operating point.

According to one aspect, it is proposed that the transition range in the partial load range is in an upper speed range, which is characterized by speeds from a transition speed. The upper speed range is in particular above the speed avoidance range, if such a range exists.

To this end, it is further proposed that the wind energy plant is characterized by a nominal speed and the transition speed is at least 80%, in particular at least 85% of the nominal speed and/or a setpoint speed of the pitch control. In this respect, it is proposed here to use the speed power control in this transition range and thus to use the speed power control for such high speeds of 80% or 85% of the nominal speed up to the full load range, i.e. up to 100% of the nominal speed. The pitch control, as far as it is active, attempts to regulate the speed to the target speed, namely by means of pitch adjustment. This target speed essentially corresponds to the nominal speed, but it may be better taken into account in the specific implementation, especially in the control structure. It is also about the transition area from the partial load range to the full load range, in which the speed setpoint is specified for the pitch control so that it provides a useful orientation. The following explanations for the nominal speed also apply to the setpoint speed.

Here it was particularly recognized that in this speed range of 80% or 85% to 100% of the nominal speed, an operating characteristic is very steep. A small change in speed therefore leads to a very large change in the generator state variable. For example, a 1% increase in speed may result in a 3% or more increase in generator state variable. Especially in the range from 95 to 100% of the nominal speed, a speed change of 1% can lead to a change in the generator state variable of 5% to 10%. Small fluctuations in the wind speed thus lead to small changes in the speed and these lead to strong changes in the generator state variable. Thus, to address these issues in this particular area, the use of speed power control there is proposed.

A speed control thus takes place, which, however, differs from the speed control in the full-load range because it uses the generator state variable, ie the generator power or the generator torque, as the manipulated variable. Due to the proposed cascade control, the rotor acceleration and thus the dynamics can also be taken into account. This in particular is not possible through the use of the operating characteristic control, because the operating characteristic control only sets the generator state variable, ie the generator power or the generator torque, for a detected speed. How quickly the speed changes is not taken into account. For low speeds, which are also usually based on low wind speeds with lower fluctuations, this can be a good and be a proven strategy. In the steep area described, in which stronger wind fluctuations are also to be expected, an improvement can be achieved by using the speed power control.

According to one aspect, it is proposed that the speed setpoint for the speed output control is specified via a transition speed characteristic curve, which forms the or a speed characteristic curve, with the transition speed characteristic curve running vertically for a speed with a speed value corresponding to the transition speed, so that the speed is constant as the generator state variable increases, until the generator state variable reaches a predetermined first generator reference value which is below a nominal value of the generator state variable.

Thus, for this transitional range, it is not only proposed to use speed power control instead of operating characteristic curve control, but another speed power curve—or speed torque curve—is also stored, which namely runs vertically from the transition speed. This also implies that when the wind speed increases, once the transition speed has been reached, the speed is initially not increased any further. The system is therefore operated there at a lower speed compared to the operating characteristic.

In this way, it can be achieved in particular that the rated speed, from which namely the full-load operation and thus the corresponding regulation begins, is not reached too early. In particular, with the operating characteristic control with the very steep operating characteristic, the speed can often reach the rated speed due to wind fluctuations. Then you have to check how to react to this, namely whether this activates the speed control, which is intended for full-load operation. There could be a frequent change between operating characteristic control and pitch control, which is avoided by the solution proposed here.

In addition or as an alternative, it can be provided that the transition speed characteristic curve still has a positive slope from the transition speed and/or that it has a positive slope from the first generator reference value, so that the values of the generator state variable increase with increasing speed until a nominal value of the generator state variable is reached. Designing the transition speed characteristic in such a way that it has a positive gradient from the transition speed and extends up to the nominal speed is not the preferred solution, but can be provided. Then you can anyway be provided that the transition speed characteristic differs in the area of the operating characteristic. In particular, it can be linear and thus lie above the operating characteristic in this area. The speed is then controlled along this transition speed characteristic, namely with the proposed speed power controller.

In particular, however, provision is made for the transition speed characteristic to run vertically from the transition speed until the generator reference value is reached, with the transition speed characteristic having a positive gradient from then on. However, it preferably reaches a nominal value of the generator state variable, ie a nominal power or a nominal torque, before the speed has reached the nominal speed. As a result, this nominal value of the generator state variable is then already reached in the part-load range and only then is the speed increased up to the nominal speed, in order then or at the same time to switch to the full-load range and the corresponding regulation in the full-load range.

According to one aspect, it is proposed that a control reserve is determined as a function of a difference between the set generator state variable and the upper generator state limit. The difference can also directly form the control reserve. This depends in particular on whether the generator state variable has the same physical unit as the control reserve. The control reserve can be specially designed as a service. If the generator state variable is in the form of a torque, a conversion of a corresponding difference in the torques into this control reserve can be considered. It is particularly advantageous if the first desired acceleration value and the detected actual acceleration value are provided in the inner cascade with the same unit as the control reserve.

It is further proposed that the control reserve is transferred from the speed power control to the pitch control. The pitch control is a speed control in which the blade angle forms the manipulated variable. In this respect, there are two speed controls, the speed power control and the pitch control, which can be matched to one another as a result.

It is further provided that the speed power control and the pitch control work at least partially in parallel and are matched to each other via the control reserve. The control reserve is an indicator of how much actuating power or actuating energy the speed power control still has available. If it has a lot of actuating energy available, the control reserve is large in terms of amount. This allows him Pitch control can be signaled that the speed power control is not yet at its limits and can still regulate.

Here it was particularly recognized that it is advantageous to initially leave the regulation to the speed power control during the transition from the partial load range to the full load range, as far as this is possible. In this case, when the wind speeds exceeding the pitch control are first activated, the rotor blades are still in an aerodynamically optimal position, i.e. they have not yet been turned out of the wind. In this case, if the pitch control becomes active, it can only mean that it puts the blades out of the wind. As a result, less power is taken from the wind. If necessary to limit assets and exposure, it should be done; if it is not yet necessary, it can still be postponed. This is exactly what can be recognized by the control reserve and this can be signaled to the pitch control.

The pitch control not only receives information as to whether the speed power control still has actuating energy, but also information about the height. This allows the transition to be continuous. If there is still a lot of actuating energy, i.e. if the control reserve is still large in terms of amount, the pitch control can remain inactive. However, if the control reserve is low but still available, the pitch control can already be active while the speed power control is also still active. Both speed controls are then active and they are matched to one another via the control reserve.

Here it was also recognized that a transition makes sense instead of waiting until the speed power control can no longer regulate due to a lack of actuating energy. This avoids hard switching and the fast controllability of the speed power control can still be used.

If the wind speed drops, the transition from pitch control to speed power control works in the same way. The operation then comes from the full load range to the partial load range and as soon as it is recognized that a control reserve is available, i.e. the generator power or the generator torque can be reduced, the transition can begin.

In principle, a switchover can also be provided, but this can be carried out at least by taking into account the control reserve to a state that is as suitable as possible or a situation that is as suitable as possible.

According to one aspect, it is proposed that the speed power control is prioritized over the pitch control, in particular in such a way that the pitch control is suppressed in whole or in part as long as the speed power control does not reach a manipulated variable limit. Preference is then given to speed power control.

In addition or as an alternative, it is proposed that the pitch control also controls the speed as a function of an actual acceleration value of the rotor and that control of the speed by the pitch control is suppressed all the more, the further the generator setpoint is below a generator setpoint limit in the speed power control.

It is therefore proposed to use the speed power control in whole or in part, as long as it can also be effective. This has already been explained above. It can also be checked whether the speed power control has reached a manipulated variable limit or not. If it reaches the manipulated variable limit, the pitch control can be switched on or its share can be increased.

According to one aspect, the pitch regulation can additionally control the speed as a function of an actual acceleration value of the rotor. This makes it possible to identify in particular whether or to what extent an excessively high speed is to be expected in the near future, which should then be counteracted by the pitch control. This counteracting can also be made dependent on how far the speed power control is still able to counteract it.

In particular, it is proposed to take into account the special structure of the proposed speed power control and to take particular account of the inner cascade there, which determines the generator setpoint as a function of the comparison between the first acceleration setpoint and the detected actual acceleration value via the second controller.

The generator setpoint thus forms a manipulated variable that can act directly or indirectly on the generator. How far this manipulated variable is from a limit, namely the generator setpoint limit, is taken into account here in order to more or less include the pitch control accordingly. If the generator setpoint is still far from the generator setpoint limit, pitch control can be severely suppressed. If the generator reference value approaches the generator reference value limit, the pitch control can be adjusted accordingly to be more strongly included in the regulation. This can be used to determine the control reserve and fed to the pitch control.

According to one aspect, it is proposed that the speed power control for controlling the speed uses the generator state variable as the manipulated variable, which is thus referred to as the generator manipulated variable, and the pitch control for controlling the speed uses the blade angle as the manipulated variable, which is thus referred to as the pitch manipulated variable, the blade angle in particular with increasing wind speed from a part-load blade angle, which is set in the part-load range, to an end angle, with the speed power control being coordinated with the pitch control, and in particular for coordinating the speed power control with the pitch control, the generator manipulated variable and the pitch manipulated variable are coordinated with one another.

Such a coordination can be such that the manipulated variable used in one controller is subtracted from the other controller, possibly with a corresponding conversion of the relevant units. However, it is also possible for the two controllers to be matched to one another in such a way that the value of the manipulated variable of one controller can be deducted directly from the other controller. To do this, the two manipulated variables would have to be coordinated with each other, e.g. via normalization. As a result, both controllers can regulate the speed without compensating for one another or leading to over-regulation, in which the manipulated variables of both controllers add up in an uncoordinated manner.

In particular, it is proposed that the manipulated variable limitation of the generator manipulated variable be taken into account and that the speed power control and the pitch control be coordinated in such a way that the speed power control, as long as it does not reach the manipulated variable limit, exerts a greater influence on the speed than the pitch control. The speed power control therefore takes precedence as long as it can implement a sufficient manipulated variable.

In addition or as an alternative, it is proposed that the manipulated pitch variable be adjusted as a function of the manipulated variable limitation of the manipulated generator variable. In particular, it is provided that a difference between the generator manipulated variable and the manipulated variable limit defines a control range of the speed power control and the pitch control variable is changed in such a way that the pitch control has less of an influence on the speed, the larger the control range of the speed power control. As a result, these two controllers can be coordinated in a simple manner. In particular, this adjustment range can be determined in a simple manner and taken into account in the coordination.

According to one aspect, it is proposed that the pitch control takes place in such a way that a second setpoint acceleration value is determined in an outer cascade from a comparison of a predefined setpoint speed with a detected actual speed via a third controller, and in an inner cascade from a comparison of the second setpoint acceleration value with a detected one acceleration setpoint, a manipulated variable, in particular the pitch manipulated variable, is determined via a fourth controller for adjusting the blade angle, and that the comparison of the second acceleration setpoint with the recorded acceleration value forms a control error and the control error is modified by means of the control reserve, and the control error modified in this way is an input variable of the fourth regulator.

The fourth controller is therefore the controller of the inner cascade of pitch control. This internal cascade takes into account an acceleration setpoint/actual value comparison. It therefore takes into account whether the detected acceleration of the rotor corresponds to the desired acceleration of the rotor. If this is not the case, there is a difference that can be described as a control error. This then leads to a corresponding reaction through the fourth controller, namely the adjustment of the blade angle. This counteracts the discrepancy between the desired and the detected acceleration.

In order to coordinate this pitch control with the speed power controller, it is now proposed that this control error, which indicates the acceleration deviation, be modified. Expressed in an illustrative manner, this control error of the inner cascade of the pitch control does not need to be corrected via a pitch control if the speed power control can achieve this via the generator instead. The speed power control can then achieve it via the generator if a corresponding control reserve is available. The dimensioning of this control reserve is preferably matched to the acceleration deviation in the pitch control. The control reserve is then simply subtracted from this acceleration control error in the pitch control. In extreme cases, when the speed power control has sufficient control reserve, the value is zero and the pitch control can be inactive.

However, it is also possible that this only reduces the control error of the acceleration in the pitch control. This then leads to the fourth controller generating or changing the pitch control variable in such a way that the pitch should be pitched, but only in a reduced size. As a result, the pitch control can be matched to the speed power control, namely in such a way that it takes into account how well or strongly the speed power control can still act at all. This is what the control reserve stands for.

The desired acceleration value, the actual acceleration value and the resulting control error of the inner cascade of the pitch control are each configured as power values. This simplifies coordination with the speed power control.

It is thus further proposed that the modification takes place in particular in such a way that the pitch change is reduced compared to when the control error is not modified, or that the pitch change is suppressed.

In addition or as an alternative, it is proposed that the modification takes place in particular in such a way that the control reserve or an equivalent, in particular proportional variable, is effectively applied to the comparison of the second acceleration setpoint with the detected acceleration value.

Such a connection is particularly useful when the pitch controller takes into account the acceleration setpoint and correspondingly the actual acceleration value, and thus also the resulting control error, as the acceleration power, as has already been described above for the speed power controller. Here, too, the acceleration power indicates the power that is required to achieve the corresponding acceleration. In this case, the control reserve can form a power value that indicates how much control power the speed power control still has available. Exactly this power value is then subtracted from the acceleration control error of the pitch control, which is then also a power.

This results in a simple control structure in which the control reserve simply needs to be generated in the speed power control and added to the pitch control, ie with a negative sign, in order to be deducted as a result. Of course, another, then probably more complicated, implementation is also possible.

According to one aspect, it is proposed that in the speed output control, the first desired acceleration value, which is determined as a function of a comparison of the specified desired speed with a detected actual speed, also as a function of a blade angle is changed. This is done in particular in such a way that the first acceleration target value is increased by an add-on value, the add-on value being determined as a function of a difference angle as the difference between the current blade angle and the part-load blade angle, in particular being determined proportionally thereto.

First of all, it should be noted that a constant part-load blade angle should be set in the part-load range. In the transition area to the full load area, however, there can be a deviation, namely due to the pitch control. This also means that the pitch control is already at least partially active. This is now taken into account by considering the difference angle, which was set by the pitch control.

This is taken into account in the speed power control when determining the acceleration setpoint. It is particularly suggested here that this is increased if such a difference angle has been determined. In particular, it is increased all the more, the larger this difference angle is, ie in terms of amount.

The consequence of this increase in the acceleration setpoint is that the speed power control attempts to compensate for this increased acceleration setpoint, ie it attempts to achieve this acceleration setpoint by appropriately setting the generator state variable, ie the generator power or the generator torque. As a result, however, the speed power control can then reach a limit. This can also be described as going into saturation. This in turn can mean that a lower control reserve or no control reserve at all is detected. This control reserve or a correspondingly lower control reserve or non-existent control reserve can react back on the pitch control with appropriate implementation and thereby activate it or increase its influence, possibly up to 100%, so that the speed power control then no longer works. This allows the speed power control to be reduced in a simple manner with active pitch control.

According to one aspect, it is proposed that a characteristic curve shift is carried out, in which the operating characteristic curve is shifted as a function of an operating point, in particular as a function of the behavior of a pitch control system, in particular as a function of a set blade angle, such that a higher value of the Generator state variable is set. If the operating characteristic is displayed in such a way that the speed is plotted on the abscissa and the generator state variable, ie the generator output or the generator torque, is plotted on the ordinate, the operating characteristic is thus shifted to the left.

The operating characteristic is intended for controlling the wind turbine in the partial load range, but the pitch control can respond during the transition to the full load range and gusty winds while the operating characteristic control is also active. At such an operating point, the operating characteristic is shifted.

Here it was particularly recognized that with such gusty winds, the wind speed can quickly fluctuate around an average wind speed. The pitch control is often not able to compensate for the rapid fluctuations, which can result in speed fluctuations. The speed can also fall below the nominal speed, which can lead to a reaction of the operating characteristic control, which namely reduces the power as the speed falls. Since the operating characteristic is particularly steep for high speeds, even small changes in speed can lead to major changes in performance. If the speed falls below the rated speed, this can lead to large reductions in power, namely to values below the rated power, although the average speed does not have to be below the rated speed.

This is avoided by shifting the operating characteristic. It was recognized that the situation described can be recognized particularly from the behavior of the pitch control. It can be derived in particular from the blade angle, which is specified by the pitch control.

It is therefore particularly proposed that the operating characteristic be shifted by a predetermined or adjustable shifting speed, which is in particular in a range from 0.3 to 1.5 rpm, preferably in a range from 0.5 to 1 rpm. With a shift of 1.5 rpm, the result is that the speed can drop by up to 1.5 rpm below the nominal speed without the operating characteristic leading to a reduction in performance.

Thus, at low speeds, such a shift has little effect, but at high speeds, when the operating curve is very steep, it has a greater effect. It also means that at high speeds near the nominal speed, the nominal value of the generator state variable, i.e. the nominal power or the nominal torque, is already reached at a slightly lower speed than the nominal speed. This can prevent a lower power than the nominal power or a lower torque than the nominal torque from being set again and again in the event of fluctuating wind speeds and correspondingly fluctuating rotational speeds despite sufficiently high wind speeds, namely in the range of the nominal wind speed.

In particular, it is proposed that the characteristic curve be shifted depending on whether the part-load blade angle is exceeded, in particular depending on an excess value by which the blade angle exceeds the part-load angle.

As long as the blade angle shows the part load angle, the pitch control has not yet become active. If it is exceeded, it has become active and a characteristic curve shift can be considered. If the partial load angle is only slightly exceeded, e.g. by only 1°, strong gusts do not have to be assumed. Therefore, the characteristic curve shift is made dependent on an overshoot value.

In particular, the characteristic curve is only shifted when the excess value reaches a minimum excess value, which is in particular in the range from 2° to 5°.

In addition or as an alternative, it can be provided that the excess value is evaluated quantitatively and the displacement speed is set as a function of the excess value.

The excess value thus enables the gustiness to be evaluated and, depending on this, the displacement speed can be selected accordingly, namely the greater the greater the excess value.

According to the invention, a wind power plant is also proposed. Such a wind energy installation has an aerodynamic rotor which is operated at a variable speed and has rotor blades whose blade angle can be adjusted. In addition, the wind power plant has a control device for controlling the wind power plant. The control device is prepared to control the wind energy installation using a method for controlling a wind energy installation according to at least one of the embodiments described above.

This means in particular that the control device can have at least one process computer on which the method is implemented. In particular, one or more operating characteristics are stored for this purpose and an operating characteristic control is implemented which, depending on the detected speed, takes a corresponding generator state variable from the relevant operating characteristic and adjusts the wind energy installation accordingly.

In particular, the control device has access to the generator of the wind energy installation and can adjust its power output and/or its generator torque. In addition, the control device has access to adjustment devices for adjusting the blade angle of the rotor blades.

In addition, a speed power control is implemented in the control device and this can receive actual values and also control the generator and the blade adjustment device.

In particular, the speed power control is implemented in such a way that it can receive both an actual speed value and a speed setpoint value and that it can also receive an actual acceleration value or calculate it from a received speed.

According to the invention, a wind farm is also proposed which has a number of wind energy installations. One, several or all of these wind turbines are designed as wind turbines according to at least one embodiment described above. In particular, the fact that several or even all of the wind energy installations of the wind farm use at least one method for controlling the wind energy installation described above means that their operation can be coordinated with one another. Particular attention should be paid to the fact that any oscillating behavior of the wind turbines can also have a mutual effect on the wind turbines. As a result, better suppression of such oscillations by the proposed methods and more constant behavior of the wind energy installations by at least one of the methods proposed above can lead to favorable behavior of the wind farm overall. In particular, a wind farm is an arrangement of wind turbines that feed into the electrical supply network via the same network connection point. With the proposed method, this feeding can also take place more uniformly and with fewer oscillations.

Figur 1

zeigt eine Windenergieanlage in einer perspektivischen Darstellung.

Figur 2

zeigt einen Windpark in einer schematischen Darstellung.

Figur 3

zeigt eine Drehzahlleistungsregelung und eine Pitchregelung jeweils in einer vereinfachten schematischen Darstellung.

Figur 4

zeigt ein Drehzahl-Leistungs-Diagramm zum Veranschaulichen des vorgeschlagenen Steuerverfahrens.

Figur 5

zeigt ein Zeit-Leistungs-Diagramm zum Veranschaulichen eines Übergangs zwischen einem oberen und einem unteren Betriebspunkt.

Figur 6

zeigt eine Drehzahl-Leistungsregelung in einer gegenüber Figur 3 erweiterten Struktur.

The invention will now be explained in more detail below by way of example with reference to the accompanying figures.

figure 1

shows a wind turbine in a perspective view.

figure 2

shows a wind farm in a schematic representation.

figure 3

shows a speed power control and a pitch control, each in a simplified schematic representation.

figure 4

shows a speed-power diagram to illustrate the proposed control method.

figure 5

shows a time-power diagram to illustrate a transition between an upper and a lower operating point.

figure 6

shows a speed-power control in a opposite figure 3 extended structure.

figure 1 shows a schematic representation of a wind turbine according to the invention. The wind turbine 100 has a tower 102 and a nacelle 104 on the tower 102 . An aerodynamic rotor 106 with three rotor blades 108 and a spinner 110 is provided on the nacelle 104 . During operation of the wind energy plant, the aerodynamic rotor 106 is set in rotary motion by the wind and thus also rotates an electrodynamic rotor or rotor of a generator, which is directly or indirectly coupled to the aerodynamic rotor 106 . The electrical generator is arranged in the nacelle 104 and generates electrical energy. The pitch angles of the rotor blades 108 can be changed by pitch motors on the rotor blade roots 108b of the respective rotor blades 108.

The wind energy installation 100 has an electrical generator 101 which is indicated in the nacelle 104 . Electrical power can be generated by means of the generator 101 . A feed unit 105 is provided for feeding in electrical power, which feed unit can be designed in particular as an inverter. A three-phase feed current and/or a three-phase feed voltage can thus be generated according to amplitude, frequency and phase, for feeding at a grid connection point PCC. This can be done directly or together with other wind turbines in a wind farm. A system controller 103 is provided for controlling the wind energy system 100 and also the feed-in unit 105 . The system controller 103 can also receive default values from the outside, in particular from a central parking computer.

figure 2 shows a wind farm 112 with, for example, three wind turbines 100, which can be the same or different. The three wind turbines 100 are therefore representative of basically any number of wind turbines in a wind farm 112.

The wind energy installations 100 provide their power, namely in particular the electricity generated, via an electrical park network 114 . The currents or powers generated by the individual wind turbines 100 are added up and a transformer 116 is usually provided, which steps up the voltage in the park in order to then feed it into the supply grid 120 at the feed point 118, which is also generally referred to as PCC. figure 2 12 is just a simplified representation of a wind farm 112, showing no controller, for example, although a controller is of course present. The park network 114 can also be designed differently, for example, in that a transformer is also present at the output of each wind energy installation 100, just to name another exemplary embodiment.

The wind farm 112 also has a central farm computer 122 . This can be connected to the wind energy installations 100 via data lines 124 or wirelessly in order to exchange data with the wind energy installations and in particular to obtain measured values from the wind energy installations 100 and to transmit control values to the wind energy installations 100 .

figure 3 shows a simplified control structure 300. It has a speed power control 302 in its upper part. It has a pitch controller 304 in the lower part. The speed power control 302 and the pitch control 304 can be coupled via a coupling connection 306 shown in dashed lines, which will be explained further below. In principle, however, both controllers can work independently of one another, so that the coupling connection 306 is only shown in dashed lines. It is provided at least according to one embodiment.

Speed power control 302 has a first summing point 308, at which a setpoint/actual value comparison between setpoint speed n _s and actual speed n _i is carried out. The resulting control error e _n is fed into the first controller 310 . The actual speed n _i is recorded at the wind energy plant and fed back via an external feedback device 312 and sent to the first summing point 308 . The first summer 308, the first controller 310 and the outer feedback 312 can be considered to be essential parts of an outer cascade of the speed power controller.

First controller 310 determines a first setpoint acceleration value a _s based on control error e _n . This is subtracted from an actual acceleration value a _i at the second summation point 314, resulting in the internal control error e _a . Thus, at the second summation point, a setpoint/actual value comparison takes place with a negative sign. Basically it is only a question of nomenclature and, to put it bluntly, is probably due to the fact that a generator torque is controlled, which brakes, i.e. counteracts acceleration. This is the reason for the chosen sign at the second summing point 314.

The resulting internal control error e _a is sent to second controller 316, which uses it to determine a desired generator value, which is used as desired generator torque or, as in FIG 3 shown variant can be designed as a power setpoint P _s . The actual acceleration value a _i is supplied to the second summation point 314 by an internal feedback 318 . The second summer 314, the second controller 316 and the inner feedback 318 can be considered as essential elements of the inner cascade of the speed power controller 302.

The acceleration values under consideration, ie the detected acceleration value a _i and the specified acceleration value a _s , are preferably taken into account as power values. In this case, an acceleration power, which is used as the value of the corresponding acceleration value, describes how much power has to be applied in order to achieve the corresponding acceleration, or how much power would be released by corresponding braking.

Nevertheless, it is necessary to use the second controller 316 to convert the inner control deviation e _a into the desired power value P _s , even if it is already a power in terms of its physical unit. Details are described below.

In any case, the desired power value P _s determined in this way is fed into the generator 320 and this has an effect on the wind energy installation illustrated in the wind energy installation block 322 . The representation of generator 320 as a separate block is essentially for illustrative purposes. In fact, of course, the generator 320 is part of the wind turbine. In particular, however, the distinction from pitch control 304 is intended to be clarified in this way.

During operation of the wind energy installation, it can therefore be operated in accordance with an operating characteristic curve control. However, at certain speeds or speed ranges, which will be explained below, the speed power control 302 shown is provided. If this is activated, it thus receives a corresponding speed setpoint. It can be fixed or part of a special characteristic.

In the case best suited for illustration, there is a constant speed setpoint n _s . The actual rotational speed fluctuates due to fluctuations in the wind speed and an external control error e _n results at the first summation point 308 .

The external control error then leads to an acceleration reference value a _s through the first controller 310 . However, this desired acceleration value is not directly converted into a generator output or a generator torque, but it is first checked how large the difference is to an existing acceleration value. The result is the internal control error e _a , from which the desired power value P _s is then determined by means of the second controller 316 and is sent to the generator 320 . As a result, the acceleration of the rotor is changed accordingly and the speed is readjusted to the specified speed.

So if a speed control is to be carried out in the partial load range, this speed power control 302 is used. The speed is thereby regulated and this also results in a corresponding output power of the generator, which is not only used for regulation, but is also output as power generated from the wind. This speed power control can be provided independently of the pitch control 304, particularly when this speed power control 302 is used to control a lower or upper avoidance speed in a speed avoidance range. Particularly for speeds below an upper speed range, that is to say not in the vicinity of a nominal speed, speed power control 302 does not need to be linked to pitch control 304, which is namely still inactive in this case.

However, a link with pitch control 304 can be considered, particularly in an upper speed range. Pitch control 304 is constructed similarly to speed power control 302 . Pitch control 304 also has an outer cascade, which has a third summing point 324 at which a setpoint/actual value comparison of the speeds is carried out in order to generate an outer control error e _n and send it to third controller 326 . External feedback 328 is also provided for actual speed n _i .

The third controller 326 then generates an acceleration reference value a _s , which is compared at the fourth summation point 330 with the actual acceleration a _i . The actual acceleration a _i is supplied to the fourth summing point 330 via the internal feedback 332 . There is an internal control error e _a .

The fifth summing point 334 can be inactive and is only required for the coordination of the speed power control 302 and the pitch control 304 . If this fifth summing point 314 is inactive, the internal control error e _a is sent directly to the fourth controller 336 . In this respect, the structure still corresponds, at least in principle, to the speed power control 302, which is why some of the designations were chosen to be identical.

However, the fourth controller 336 then generates a desired value for the blade angle α _s , which is fed to a pitch system 338 accordingly. The blades are thus adjusted by means of the pitch system 338 in such a way that the actual acceleration value is adjusted to the setpoint acceleration value in order to also achieve an adjustment of the actual speed and the setpoint speed. In this respect, the pitch system acts on the wind turbine block 340, which could correspond to the wind turbine block 322. Here, too, the pitch system 338 is of course part of the wind energy installation, and the entire regulation can also be regarded as part of the wind energy installation. In this respect, the wind turbine block 340 also serves to illustrate that the pitch system 338 acts on the wind turbine. The pitch regulation thus acts on the wind turbine by specifying the target angle α _s .

Despite the similar structures of speed power control 302 and pitch control 304, the two controls use different manipulated variables. Speed power control 302 uses a generator power or a generator torque as a manipulated variable, whereas pitch control 304 uses a blade adjustment as a manipulated variable. This is of course also taken into account in the corresponding controllers, in particular in the second and fourth controllers 316, 336, respectively.

However, the two structures are selected to be similar in that the target/actual value comparison in the second summing element 314 and in the fourth summing element 330 can be the same to the extent that they can be based on the same physical unit for the acceleration. In particular, it is suggested that in both cases the acceleration values should be considered as power. This makes it possible to couple or coordinate the speed power control 302 with the pitch control 304 .

The second controller 316 determines a control reserve P _R for this and outputs it. The control reserve P _R states how much power is available to the speed power controller 302 for controlling. For example, the wind turbine already gives maximum power off, the reserve power would be zero, to name an extreme case. The value zero is then present at the fifth summation point 334, ie it does not change the pitch control at all. The pitch control would then adjust the rotor blades out of the wind quite normally to reduce the rotor acceleration. At the same time, the speed power control 302 would not be able to bring about any changes, since its manipulated variable, namely the generator power (which could also be the generator torque), is already at a limit. which can also be referred to as saturation. In this case, in fact, only pitch control 304 would still be active, and that 100%, whereas speed power control 302 would no longer be active. However, this could change quickly if the wind drops, for example.

However, if there is a control reserve, this is subtracted from the setpoint/actual value difference at the fifth summing point 334 . This internal control error e _a is therefore the rotor acceleration that remains after the setpoint/actual value comparison and would have to be corrected. To this end, it is fed into the fourth regulator 336 . However, if speed power control 302 still has sufficient control reserve available, this is subtracted from the acceleration to be controlled. In extreme cases, this can mean that this is deducted to zero. In this extreme case, only speed power control 302 would be active, but pitch control 304 would not.

Of course, it can also be the case that the control reserve P _R is significantly greater than the output of the fourth summing point 330 . The output of the fifth summing junction 334 would then go negative. However, due to limitations, this would not lead to a blade angle or pitch adjustment pointing in the other direction. Here it is particularly important to take into account that the rotational speed power control 302 and the pitch control 304 can work in an overlapping manner, particularly in the transition range from the partial load range to the full load range. At the very beginning of the full load range, however, the rotor blades are still optimally positioned in the wind. The pitch control can then only turn the rotor blades in one direction. A negative output value at the fifth summing point 334 would have no effect in this case. On the other hand, if the rotor blades were already turned slightly into the wind, a negative output value at the fifth summing junction 334 could result in the blades being fully turned into the wind again.

These were just a few illustrative examples to explain the basic principle. The detailed control steps may of course depend on other details, including integral terms, particularly in the second controller.

figure 4 shows a speed-power diagram in which the generator power P is plotted against the speed n of the rotor. Beginning with the starting speed n ₀ , a speed-power characteristic curve is shown which rises in a strictly monotonous manner until the rated speed n _N is reached at the rated power P _N and forms an operating characteristic curve 402 . This operating characteristic is interrupted in a speed avoidance area 404, which is indicated there by the dotted line. In addition, it is proposed that in an upper speed range 406 the operating characteristic should also not be used to control the wind energy installation, so that it is only shown there as a broken line.

The speed avoidance region 404 basically encloses a resonance point. At this resonance point, a resonance frequency of the wind turbine can be excited by the resonance speed n _R that is drawn in. Therefore, the wind turbine should not be operated at this speed if possible.

When the wind speed increases, the speed increases gradually according to the operating characteristic curve 402, starting from the starting speed n ₀ , depending on the increase in the wind speed. Depending on which speed is set, an assigned power is set according to the operating characteristic 402 .

If the speed reaches the lower avoidance speed n ₁ , a speed power control is switched on, as is known as speed power control 302 in 3 is shown. The coupling connection 306 is not relevant here and can either be omitted or will have no effect. If the wind speed continues to increase, this would lead to an increase in the speed, but the speed power control counteracts this. She works like related to 3 explained, wherein the lower avoidance speed n ₁ can be entered as a target speed n _s , so it is then given to the first summing point 308.

As a result, power increases to counteract the acceleration of the rotor. This can be implemented particularly well by the proposed cascade control, which determines an acceleration reference value in the outer cascade and regulates this in the inner cascade. This results in the illustrated left branch 408. The higher the wind speed then becomes, the higher this vertical branch 408 rises.

It can then be provided to switch from the left branch 408 to the right branch 410 at the upper avoidance speed n ₂ . If the wind speed then continues to increase, the operating point moves further up on the right branch 410 until the operating characteristic 402 is reached. If the wind speed then increases further, the working point on the operating characteristic 402 moves further in the direction of higher speeds and also higher power.

Switching from the left to the right branch, or vice versa when the wind speed drops, can depend on limit values. In particular, it is proposed that there is a change from the left branch 408 to the right branch 410 when the generator state variable, in this case the power P, has reached an upper limit value. Conversely, a change, especially when the wind speeds are falling, can be provided if the system is being operated with an operating point on the right branch 410 and has reached a lower power limit. It can be specifically provided that a change has to be made when the upper limit is reached on the left branch or the lower limit is reached on the right branch.

However, it can be provided that there is also the possibility of changing beforehand. For this, an aerodynamic efficiency of the current working point can be determined. This depends in particular on the speed factor, i.e. the quotient of a blade tip speed divided by the current wind speed. In the partial load range, the wind energy installation is usually designed in such a way that the speed ratio is as ideal as possible. This design usually leads to the operating characteristic curve used, that is to say operating characteristic curve 402 in this case. If the operating point lies on this operating characteristic curve 402, it is aerodynamically optimal. This results in an optimal aerodynamic efficiency.

If the speed is now maintained, for example, at the lower avoidance speed n ₁ , with the power increasing according to the left-hand branch 408, then this operating characteristic curve 402 is exited and thus also the aerodynamically optimal operation. The aerodynamic efficiency thus decreases. It can also be calculated, since a characteristic map of the wind energy installation and thus the efficiencies of various operating points of a wind energy installation are usually well known. From the speed recorded and the power generated, the efficiency and from this, via the known map, the wind speed can also be derived.

If the wind speed was derived, it can also be calculated which operating point the wind energy installation would assume after a change to the right-hand branch 410 . This can also be carried out using the known map from which the aerodynamic efficiency can then be calculated and thus the aerodynamic efficiency of the current operating point, i.e. according to the example mentioned on the left branch 408, can be compared with the efficiency of the operating point on the right branch 410, which the wind energy plant assumes after a change would. If the efficiency of the operating point, which was calculated for after a change, is higher than the current aerodynamic efficiency, a change can be considered. However, the change does not have to be carried out immediately if, for example, the difference between the two efficiencies is small and it is not yet foreseeable whether the wind speed will continue to rise in order to stick with the example above.

If, with falling wind speed, an operating point moves from above towards the upper avoidance speed n ₂ , the method can be used analogously to change from the right branch 410 to the left branch 408 .

A switching operation can be performed as shown in figure 5 is described. figure 5 is an illustration of a change using an example in which the wind speed drops and the power that can be generated P drops as a result. This is illustrated schematically by the strictly monotonously falling dashed line. It is assumed that the upper avoidance speed n ₂ is reached at time t ₁ coming from above. Then the speed power control starts and tries to keep the speed at the upper avoidance speed n ₂ . This leads to a decrease in power, which is indicated by the power control branch 550 . The dashed power curve 552 is therefore above the power control branch 550. As a result, less power is converted, as a result of which the rotor is braked less or no longer. The speed can be maintained.

At the start time t ₂ , namely at the start of a change process, the power has dropped so much that it has either already reached a lower limit value or that the aerodynamic efficiency of the working point has meanwhile become so poor that a change is advisable. The power control branch 550 basically corresponds to the right branch 410 of FIG 4 . In 4 However, the power decrease is shown over the speed n, so that the right-hand branch runs vertically. In figure 5 however, the course of the power control branch 550 is shown over time, so that it is not vertical.

If a change is now initiated, the power increases sharply, as can be seen in the rise branch 554 of the time profile shown. Another upper branch 556 and a Drop branch 558 together with the rise branch 554 form a time course for the generator state variable to be set, namely the power to be set here.

At the time t ₃ after the changeover, the changeover is complete, so that the operating point is then on the left branch 408, but at the foot of the left branch 408, because in the representation of FIG figure 5 the waste branch 558 ends on the dashed power curve 552. The optimal power that can be generated is then delivered, so that the optimal operating point has been set, which in turn lies on the operating characteristic. However, the optimal operating point, which can also generally be referred to synonymously as the operating point, does not have to be reached when changing.

Here, in particular, the time difference between the start time t ₂ and the post-change time t ₃ can be designated and specified as the change time T _w . The distance between the power level of the upper branch 556 and the power achieved after the change can be referred to as the alternating power P _w . The rise branch 554, the upper branch 556 and the fall branch 558 form approximately a trapezium and to calculate the alternating energy, the product of the alternating time T _w and the alternating power P _w can be used for simplification. In order to take into account the slopes of the rise branch 554 and the fall branch 558, the time that one of the two edges needs can be subtracted as the change time T _w instead of the difference between the start time t ₂ and the post-change time t ₃ , i.e. for example the time that the Drop leg 558 needed to drop from the top leg 556 to the dashed power curve 552.

Correspondingly, when the wind speed increases, it is possible to switch from the left branch 408 to the right branch 410, in that the power falls below the ideal power curve, in order to thereby allow the rotor to be accelerated.

In the 4 the use of a speed power control for the upper speed range 406 is also explained. Accordingly, it is provided that from the transition speed n ₃ the speed power control is used, as in 3 is explained. In particular, the coupling between the rotational speed power control 302 and the pitch control 304 can also be provided here by means of the coupling connection 306 .

To this end, it is proposed that the rotational speed power control be given a rotational speed according to a transitional rotational speed characteristic curve as the rotational speed setpoint value. Such a transition speed characteristic 412 has a vertical branch 414 and a residual branch 416, which is on connects the vertical branch 414. The rest branch 416 is followed by a horizontal branch 418 which has a nominal power P _N and reaches a nominal speed n _N and can reach beyond it. The horizontal branch 418 can also be viewed as part of the transition speed characteristic curve 412 .

The vertical branch 414 thus means that when the wind speed increases, the transition speed n ₃ initially forms the setpoint speed for the speed power control. If, with increasing wind speed, the power reaches a reference power value P _Ref which is below the nominal power P _N , the restast 416 is used. When the wind speed increases, the restaur 416 should finally lead the operating point up to the nominal power P _N . The nominal power P _N should preferably be reached before the nominal speed n _N is reached.

For this purpose, the Restast 416 can have a positive gradient and, as in 4 shown to be straight. According to the rest 416, the power then increases proportionally with increasing speed. However, other courses are also conceivable, e.g. B. according to a second order polynomial, so that the _Restast 416 can then be curved to thereby achieve the nominal power PN.

In any case, such a rest 416 reflects a connection between speed and power. This relationship can be used in such a way that the operating point is identified on the basis of the output power and the associated speed is then entered as the setpoint speed in the speed power control. As a result, the wind energy installation can also be guided along this residual button 416 by means of the speed power control.

Alternatively, a negative gradient can also be provided for the vertical branch 414, so that the speed decreases somewhat as the power increases. This is indicated by the alternative branch 420, which is shown in dashed lines. Its gradient can be up to 10%, according to which the speed decreases by up to 10% of the nominal speed when the power increases by the nominal power. This maximum gradient of 10% in terms of amount therefore relates to the change in speed per change in power, in each case based on the nominal speed or nominal power. The regulation of the alternative branch 420 can also take place in that the power is detected and, depending on this, the speed assigned according to the alternative branch 420 is used as the setpoint speed for the speed power control.

Particularly in the area of the residual button 416, it is possible that the speed power control, as described in 3 explained above, no longer has sufficient control reserve. Accordingly, again referring to 3 , the coupling connection 306 transmits a correspondingly small value to the pitch controller 304, which is taken into account there at the fifth summing point 334. Accordingly, pitch control 304 can then already become active.

figure 6 shows a speed power control according to the upper part of the 3 , but with further elements, partly with more details and partly in a slightly different representation. In any case, the speed power control 602 of 6 also a first summing point 608, which basically carries out a setpoint/actual value comparison for the speed and thus generates an external control error e _n for the speed. This is entered into a first controller 610, which however has further details here and is therefore only shown as a dashed block. First controller 610 also outputs a target acceleration a _s . For this purpose, a setpoint/actual value comparison takes place at the second summation point 614 and is input into a second controller 616 . This in turn outputs a target generator status variable, namely a target power P _S in this case. An outer return 612 and an inner return 618 are also provided.

In this respect, the function also corresponds to that 3 for speed power control 302 explained functionality. The setpoint power P _S output from the second regulator 616 is given to the wind energy plant block 622, which here, however, contains the generator. In addition, the wind turbine block 622 outputs a speed n and this is passed through a filter block 650 . The filter block 650 thus outputs two filtered speeds, with the filtering being able to be different. For the internal feedback 618, the rotational speed in the acceleration block 652 is converted into the detected acceleration a _i by derivation, among other things. Such a differentiation is not provided in the outer feedback 612 and it can therefore make sense for the filter block 650 to be filtered differently for the inner feedback 618 than for the outer feedback 612.

The first controller 610 is constructed in such a way that the speed control error is converted into a provisional acceleration value a _v in the speed conversion block 654 . This is converted into the desired acceleration value a _s via an acceleration limitation block 656 . The provisional acceleration value a _v is already provided as a performance. An equivalent pitch power P _α can be subtracted from it, namely in the third summing point 658. The equivalent pitch power P _α is determined by the pitch power block 660. The difference between an existing blade angle α and a minimum blade angle α _min is taken into account in the blade angle performance block 660 with the aid of the fourth summing point 662 . In addition, information about the current pitch control Pit_C is taken into account.

This difference between the current and minimum blade angle is particularly important. If the current blade angle α corresponds to the minimum blade angle α _min , the difference is zero and there is no need to take it into account. The output value of the blade angle power block 660, ie the equivalent blade angle power P _α , can then be zero. However, if the blade angle is misaligned with respect to the minimum blade angle, this means that the power consumption is no longer optimal, in particular that it is reduced, namely by this equivalent blade angle power value P _α . This is therefore taken into account at the third summing point 658 in the first controller 610 . The provisional acceleration a _v , which is intended to lead to the target acceleration a _s , can thus be reduced.

The stronger this connection is selected, the sharper the speed control described is decoupled from the pitch control by means of the generator power.

In special operating situations in which it is natural to exceed the minimum blade angle, for example during the system start-up process, deactivation of the activation described is proposed.

The provisional acceleration value a _v is modified accordingly and the result is also given via the acceleration limitation block 656, so that the desired acceleration value a _s results.

The setpoint/actual value comparison of the acceleration values then leads to the internal control error e _a , which is input into the second controller 616 in the acceleration conversion block 664 . Particular consideration is given to a time constant in the acceleration conversion block 664, which may be referred to as reset time. The result is then fed to the integrator 666. The output of the integrator 666 is basically the setpoint power P _s to be determined, with this being initially given via a power limitation block 668 in order to take limitations into account.

In order to prevent the integrator 666 from further integrating although a limit has already been reached in the power limit block 668, a comparison is made between the unlimited and the limited power in the fifth summing point 670. If the limit in the power limit block 668 does not apply, the result has the difference formation in the fifth summing point 670 has the value zero. However, if the limit occurs, the difference is fed back to the input of the integrator 666 via the integrator limit block 672 by subtracting it from the result of the acceleration conversion block 664 at the sixth summing point 674 .

Furthermore, in the speed power control 602 of 6 a characteristic shift is provided. The speed power characteristic can be shifted by the speed shift value n _k to be taken into account in the speed power control. For this purpose, the rotational speed shift value n _k is added to the actual rotational speed value at the seventh summing point 676 . The speed shift value n _k can be 0.3 to 1.5 rpm, for example. This value is added to the measured or recorded speed n, which means that, from the point of view of the speed power controller, the actual speed is slightly higher than it actually is. This leads to the effect that the operating characteristic, ie the speed power characteristic, is shifted to the left by the corresponding value of the displacement speed n _k .

The activation of the shift and also its size can depend on a set blade angle, and thus on a behavior of a pitch control, such as the pitch control 304 of FIG figure 3 . According to the pitch control, the target value for the blade angle α _s can be used as a criterion for the displacement. In this way it can be achieved that in the event of gusts, a drop in speed does not lead to a reduction in power if this drop in speed is less than the displacement speed below the nominal speed.

The invention thus serves to control the speed of the wind turbines in the partial load range, i.e. as long as the speed is controlled to a large extent by the system output or the generator torque, and not yet primarily by the blade angle.

In particular, an improvement in the power curve of a wind energy plant and a circumvention of speeds can be achieved, above all to avoid operation in resonance points. This means that loads and tonality can be avoided, to name just two examples.

In particular, the following is proposed:
A cascaded control structure is proposed, with an outer loop for speed control and an inner loop for acceleration control. This enables an advantageous coupling to a full-load controller, namely in particular to the pitch controller.

A transition characteristic is proposed, via which the proposed control concept, in particular the speed power control, can be made compatible with the pitch control.

Transition functions or concepts, which can also be referred to as transition functions, for the controlled passage through speed avoidance areas of resonance points, as well as the selection of the switching times for passing through the speed avoidance areas, are proposed.

At least one section with a non-vertical speed characteristic is proposed for sections in the partial load range in which speed power control is proposed, especially in the upper speed range, especially as a linear section with a constant slope. The section thus specifies power-dependent speed setpoints.

Claims (21)

A method for controlling a wind power plant (100), the wind power plant (100) having an aerodynamic rotor (106) which is operated at a variable speed and the rotor blades (108) which can be adjusted in their blade angle, wherein - the wind turbine (100) is controlled in a partial load range by an operating characteristic control which uses an operating characteristic, wherein - the operating characteristic (402) specifies a relationship between the rotational speed and a generator state variable to be set, the generator state variable to be set being a generator output or a generator torque, and - the operating characteristics are controlled in such a way that a value of the generator state variable that is predetermined by the operating characteristics is set as a function of a detected rotational speed, and - the wind energy plant is controlled in a full-load range by a pitch controller (304), in which the speed is controlled by adjusting the blade angle to a speed setpoint, wherein - In at least one predefinable speed range of the part-load range and/or in a transition range from the part-load range to the full-load range, the wind turbine is controlled by a speed power controller (302), in which the speed is controlled by adjusting the generator state variable to a speed setpoint, wherein - the speed power control takes place in such a way that - In an outer cascade, a first setpoint acceleration value (a _s ) of the rotor is determined from a comparison of a predefined setpoint speed with a detected actual speed via a first controller (310), - In an inner cascade from a comparison of the first acceleration target value with a detected actual acceleration value of the rotor via a second controller (316) a generator setpoint is determined as a setpoint for the generator state variable. Method according to claim 1, characterized in that - the speed power control (302) has an integral component, in particular that - The second controller has the integral part. Method according to claim 1 or 2, characterized in that - in the speed power control - The first desired acceleration value and the actual acceleration value are each designed as an acceleration power, where - the acceleration power is associated with a rotor acceleration and describes a power required to produce the rotor acceleration. Method according to one of the preceding claims, characterized in that - in the speed power control - The outer cascade, in particular the first controller, has at least one acceleration limit value for determining the first desired acceleration value, and preferably - the at least one acceleration limit value can be set and/or - an upper and a lower acceleration limit with different values are provided as the at least one acceleration limit, and or - The inner cascade, in particular the second controller, for determining a variable or the manipulated variable for setting the generator state variable - has an integral term with an integrator limitation, and optional - the integrator limitation is adjustable and/or - has different upper and lower limit values. Method according to one of the preceding claims, characterized in that - the at least one predefinable speed range is provided as a speed avoidance range, which is characterized by a lower avoidance speed and an upper avoidance speed, and - the speed power control is used when the lower avoidance speed is reached with increasing speed, or the upper avoidance speed is reached with falling speed, where
when using the speed power control, the wind turbine as a function of a prevailing wind speed - Can be operated at a lower operating point in a lower speed range, which includes the lower avoidance speed, and - Can be operated at an upper operating point in an upper speed range, which includes the upper avoidance speed, and - A change between the lower and upper working point is carried out depending on an aerodynamic evaluation of the lower and upper working point. Method according to one of the preceding claims, characterized in that - An at least two-stage test is carried out to switch between one or the lower and upper working point, wherein - In a first stage, it is checked whether the working point to which the change is to be made has a higher aerodynamic efficiency than the working point from which the change is to be made, and - It is checked in a second stage whether the generator state variable, which is determined with the part-load speed control, reaches an upper or lower generator state limit, in particular - Depending on the test in the first stage, a change is allowed, and depending on the test in the second stage, a change is prescribed. Method according to one of the preceding claims, characterized in that - A change between one or the lower and upper working point is carried out if - the operating point to which the change is to be made, - for a predeterminable test period, - Has a higher aerodynamic efficiency than the operating point from which to change, preferably - the predeterminable test period is in a range from 5 seconds to 10 minutes, in particular in the range from 10 seconds to 5 minutes. Method according to one of the preceding claims, characterized in that - for speed power control, the speed setpoint - is specified as a constant, or - Is specified via at least one speed characteristic, the speed characteristic each having speed values as a function of the generator state variable to be set, and in particular the speed characteristic curve is designed in such a way that it has a negative slope, at least in sections, so that the speed values decrease with increasing values of the generator state variable. Method according to one of the preceding claims, characterized in that - to switch between one or the lower and upper working point - a changeover time is specified, which is preferably less than 20 seconds, in particular less than 10 seconds, and/or - A time profile for the generator state variable to be set is specified in order to thereby control the change. Method according to one of the preceding claims, characterized in that - The transition range in the partial load range is in an upper speed range, which is characterized by speeds from a transition speed, wherein the upper speed range is in particular above a speed avoidance range, wherein - the wind turbine is characterized by a nominal speed, and the transition speed is at least 80%, in particular at least 85% of the nominal speed and/or a target speed of the pitch control. Method according to one of the preceding claims, characterized in that - Is specified for the speed power control of the speed setpoint via a transition speed characteristic, which forms the or a speed characteristic, wherein - for a speed with a speed value corresponding to the transition speed, the transition speed characteristic curve runs vertically, so that the speed is constant as the generator state variable increases, until the generator state variable reaches a predetermined first generator reference value, which is below a nominal value of the generator state variable, and/or - the transition speed characteristic has a positive gradient from the transition speed and/or from the first generator reference value, so that the values of the generator state variable increase with increasing speed until a nominal value of the generator state variable is reached. Method according to one of the preceding claims, characterized in that - a control reserve is determined as a function of a difference between the set generator state variable and one or the upper generator state limit, or the difference is the control reserve, - the control reserve is transferred from the speed power control to the pitch control, and - the speed power control and the pitch control work at least partially in parallel and are matched to one another via the control reserve, and/or - a changeover or a transition between the speed power control and the pitch control takes place depending on the control reserve. Method according to one of the preceding claims, characterized in that - the speed power control is prioritized over the pitch control, in particular in such a way that the pitch control is suppressed in whole or in part as long as the speed power control does not reach a manipulated variable limit, and/or - the pitch control controls the speed additionally as a function of an actual acceleration value of the rotor, and control of the speed by the pitch control is suppressed the more the further the generator reference value lies below a generator reference value limit in the speed power control. Method according to one of the preceding claims, characterized in that - the speed power control for controlling the speed uses the generator state variable as the manipulated variable, which is referred to as the generator manipulated variable, and - the pitch control for controlling the speed uses the blade angle as a manipulated variable, which is referred to as the pitch manipulated variable, the blade angles being increased in particular with increasing wind speed from a part-load blade angle that is set in the part-load range to an end angle, where - the speed power control is coordinated with the pitch control, and in particular - To coordinate the speed power control with the pitch control, the generator manipulated variable and the pitch manipulated variable are coordinated with one another. Method according to claim 14, characterized in that - one or the manipulated variable limitation of the generator manipulated variable is taken into account, and - the speed power control and the pitch control are coordinated in such a way that the speed power control, as long as it does not reach the manipulated variable limit, exerts a greater influence on the speed than the pitch control, and/or - The pitch control variable is set as a function of the control variable limitation of the generator control variable, in particular - a difference between the generator manipulated variable and the manipulated variable limitation defines a control range of the speed power control, and - the pitch control variable is changed in such a way that the pitch control has less influence on the speed, the larger the control range of the speed power control. Method according to one of the preceding claims, characterized in that - the pitch control takes place in such a way that - a second acceleration setpoint is determined in an outer cascade from a comparison of a specified setpoint speed with a detected actual speed via a third controller, - a manipulated variable, in particular a pitch manipulated variable, for adjusting the blade angle is determined in an internal cascade from a comparison of the second acceleration setpoint with the detected actual acceleration value via a fourth controller, and - The comparison of the second acceleration target value with the detected actual acceleration value forms a control error and the control error is modified by means of a control reserve, and the control error modified in this way forms an input variable of the fourth controller, with preferably - the desired acceleration value, the actual acceleration value and the resulting control error are each in the form of power values, and/or - the modification takes place in particular in such a way that the pitch change is reduced compared to when the control error is not modified, or that the pitch change is suppressed, and/or wherein - The modification is carried out in particular in such a way that the control reserve or an equivalent, in particular proportional variable, is effectively applied to the comparison of the second desired acceleration value with the detected actual acceleration value. Method according to one of the preceding claims, characterized in that - In the speed power control, the first acceleration setpoint, which is determined as a function of a comparison of a specified setpoint speed with a detected actual speed, is also changed as a function of a blade angle, in particular that - The first acceleration target value is increased by an added value, the added value being determined as a function of a difference angle as the difference between the current blade angle and a blade angle or the partial load blade angle, in particular being determined proportionally thereto. Method according to one of the preceding claims, characterized in that - a characteristic curve shift is carried out, in which the operating characteristic curve is shifted as a function of an operating point, in particular as a function of a pitch control behavior, in particular as a function of a set blade angle, such that a higher value of the generator state variable is set at the same speed, in particular in such a way , that - The operating characteristic is shifted by a predetermined or adjustable displacement speed, which is in particular in a range from 0.3 to 1.5 rpm, preferably in a range from 0.5 to 1 rpm. Method according to claim 18, characterized in that - the shift in the characteristic curve - is carried out depending on the partial load blade angle being exceeded, in particular - depending on an excess value by which the blade angle exceeds the part load angle. Wind turbine (100) with - an aerodynamic rotor (106) which is operated at a variable speed and which has rotor blades (108) with an adjustable blade angle, and - A control device for controlling the wind turbine (100), wherein - The control device is prepared to control the wind turbine by means of a method according to one of the preceding claims. Wind farm (112) with a plurality of wind energy installations (100), one, several or all of the wind energy installations being designed as wind energy installations according to claim 20.

Images